Against the illusion theory of temp Title (Proceedings of the CAPE Internatio II: The CAPE International Conferen Philosophy of Time ) Author(s) Braddon-Mitchell, David Citation CAPE Studies in Applied Philosophy 2: 211-222 Issue Date 2014-03-24 URL http://dx.doi.org/10.17983/203255 Right Type Departmental Bulletin Paper Textversion publisher Kyoto University
Against the illusion theory of temporal phenomenology David Braddon-Mitchell Introduction This is a paper which tackles not so much the many arguments for the A theory of time, so much as its motivation. Over the last half century, arguments for and against the A theory have raged almost without cease. The present author thinks that the battles have largely been won by the B theorist, but nothing hangs on that for our present purposes. What is striking, though, is that whenever there is what seems (to some of us) a decisive response to the A-theorist, A theorists seem not to quaver for a moment. Something tells them that their view must be true, even if they can t quite see why. Sometimes explicitly as an argument, sometimes tacitly, this is because it just seems to many as though the A theory is true. Or so they believe. I intend this in absolutely literal way. The phenomenology associated with perceiving the world is taken as representing the present as special in some way. It might be that it seems as though the present is special in some way (other than being what is present to each time-slice), or that there is a special perceived quality of flow, or that the past is unreal, or that the future is open, or that there is a unique frisson of being and becoming that shows the world to be an A-world. In this paper I ll just pick on one of these: the perception of flow. I hope and believe that nothing hangs on this choice. So the underlying motivation for the A theory, and perhaps a powerful argument in its own right goes something like this. 211
212 David Braddon-Mitchell A-Motivation (1) Phenomenology of perception represents the world as possessing A-properties (e.g. Flow). (2) Absent powerful arguments to the contrary, the deliverances of perception should be taken seriously as providing powerful prima facie evidence for the veridicality of how they represent the world. (3) There are no such arguments to the contrary. Conclusion: The world likely possesses A-properties. Most B theorist seem to accept not just the validity of this argument, and the truth of its first two premises. In which case they have to accept that the world does seem as though it contains A-properties, and must therefore be what I ll call Illusionists: they must hold that there is an illusion that there are A-properties. Thus they in effect deny premise 3 of A-Motivation. Recently Laurie Paul (Paul 2010) in an article in the Journal of Philosophy argues that there must indeed be such an illusion, and endeavours to explain what causes it. Simon Prosser (Prosser 2013) Brad Skow (Skow 2011) and Jack Smart (Smart 2006 and many earlier pieces) are further examples of B theorists who argue in this general way. Arguing in such a way gives a lot of ground to the A theorist; ground that it is the object of this paper to recapture. Instead this paper focusses on the first premise: that phenomenology has a representational content, and that its content is of perception and argues against it. Denying that is denying what the illusionist believes. It s not that things seem as though there are A properties, it s just that some people mistakenly believe that they do. I will deny a doctrine I call ESP: the Evidential Significance of Phenomenology. The conclusion then is that ESP is false. Method In fact there are many things which we might deny which would undermine premise 1, and here are some of them: (1) The phenomenologies exist and have (intrinsic) content. (2) The content is of flow. (3) The content is unambiguously and unconditionally of flow.
Against the illusion theory of temporal phenomenology 213 (4) The content is on a person to person case consistently about flow, regardless of theory. (5) We have reliable access to that content (perhaps given that it s phenomenal via revelation) independent of best metaphysical argument on other grounds. If phenomenologies don t have content premise 1 is straightforwardly false. If it has content but it s not as of A properties being instantiated it s false. If it sometimes has such content but not consistently so between people or times, or is ambiguous in some way, then Premise 1 does not support the conclusion. The structure of the paper will then be straightforward. I ll review a wide range of views in the philosophy of mind about representation, and argue that on all of them one of these things is false, and thus that either premise 1 is false or fails to support the conclusion, and thus that ESP is false. Of course this does not logically rule out that ESP is true. 1-5 above is a conjunction of things which if correct would vindicate it. But I think of my argument as being one directed at someone uncommitted in the philosophy of mind and time. The idea is that the amount of logical space occupied by the conjunction of all of these views is very small. What are the odds that that uncommitted philosopher should end up there on reflection? It s not just that these views on representation that I ll discuss are incompatible with ESP. There s also a piece of data to accommodate that the believer in ESP had better accept. And that s that things are at the very least not as simple as they seem. After all, to many of us it does not seem as though our phenomenologies represent A-properties as obtaining. I for one am a B theorist but I don t think there s anything illusory about my phenomenology. It makes no claims about A-properties at all as far as I can tell. And there are plenty others who hold the same (informal surveys of beginning philosophy of time students say about 80%; but I put little store in that). If I accept the good faith of the A-theorist, and she of the B-theorist, then we need to add something else to our list of things that the right theory of representation needs to account for: API, or Apparent Phenomenal Inconsistency. We don t all take our phenomenology to represent the world as being inhabited by A-properties, so either it does for some of us and not for others, or some of us are mistaken about the content of our representations.
214 David Braddon-Mitchell No intrinsic representation The first set of positions I want to consider are the ones according to which phenomenology has no intrinsic representational power. By this I mean that if there are states that are intrinsically phenomenological, the representational power they have (if any) comes not from their intrinsic nature, but because they are associated with another state which does have representational power. This could go in at least two ways. According to one, the phenomenal state has no representational power at all, but covaries with a functional state which does. An example: suppose that the state which is the phenomenology of redness occurs in the brain along with a state which makes us behave as though objects have a certain reflectance profile. The second state might be doing the representing, the first just as it were along for the ride. This view makes particularly good sense of inverted spectra, since in that case it s just different representational states combined with the same phenomenal states. The other way this might work is that the phenomenal state might be representation but not intrinsically: it might be that it is has a certain causal history, or does certain functional work, and it s in virtue of that it has representational power. Tokened alone, without its causal history or functional powers, it would be non representational. These are two of the most attractive options for functionalists about the mind, and thus take up a good chunk of logical space. If either is true, then the best explanation of the difference between those who see flow and those who do not might not be that there is any difference in phenomenology: rather it s a difference in the representational states that co-vary with them. Now one imagines that the difference between the A-theory and the B-theory are not ones which affect ordinary behaviour, and thus a invisible to certain kinds of functional analysis. But of course A-theorists and B-theorists are inclined to make different sorts of verbal judgements, and to perform different inferences when using temporal concepts and vocabulary. And it s at that functional grain that I think we can locate the difference in representation. But if that s so it explains API straightforwardly. When an A theorist tokens some temporal phenomenology, she is inclined to reason and judge in A theoretic ways. When a B theorist tokens some temporal phenomenology, she is inclined to reason and judge in B theoretic ways. But these states of being inclined to reason and judge just are the representational states. So it is true that the A theorist has states that represent A properties:
Against the illusion theory of temporal phenomenology 215 but only because she is an A theorist. And similarly the B theorist has states which do not so represent, but only because she is a B theorist. So the representations are not evidence for the truth of the theories, because they are the result of (perhaps) belief in the theory. And so, on these views, there is no ESP no evidential significance to phenomenology. Uniform but ambiguous or abstract representation Suppose for a moment that there are phenomenologies of time which do have intrinsic representational power. This would not be enough to vindicate the first premise of A-motivation. For there wold still be the question of at what exact level of abstraction this representation is to be found. Abstract representations occur all the time. People may agree that something seems unjust. But they may disagree in the end about what justice is. The representation may be an abstract one that captures something like a very broad role that is in common between many concepts of justice that differ on the details. But of course its perfectly possible that the more convinced someone is of the rightness of their own view about the nature of justice, the more they will assume that the representational content of the phenomenology shares the detail that their own theory of justice does. The more they are unacquainted with alternative views, or perhaps don t take them seriously, the more likely this would be. Something like this could be going on in the case of temporal phenomenology. Perhaps, our phenomenology represents the world as changing, for example. But the exact analysis of change is a tricky philosophical matter. The A theorist might be inclined to argue as follows: (1) Our phenomenology represents the world as containing change (2) True change involves A-properties (3) [from (1) and (2)] Our phenomenology represents the world as containing something (as a matter of fact) involves A properties (4) Phenomenology provides a good prima facie guide to the world (5) The world contains A-properties. But this is not a good argument. There are two ways to show this. One is just to note that (2) pretty much contains the conclusion since it s a background assumption for all sides that there is change, and (2 )would of course be disputed. The other is to note that its not enough
216 David Braddon-Mitchell for the argument to show that according to the metaphysical lights of the A theorist our phenomenology represent something which as a matter of fact involves A properties. An argument from representation requires that our phenomenology represents it as involving A-properties. So here then is another way in which it can be theory that drives our views about representation, rather than the other way around. In the previous section theory determined what the content of representation was, and thus the content of representation was no guide to the truth of the theory. In this section theory drives beliefs about the content of representations, where A theorists and some B theorists come to have opinions about the content of representation which are more fine-grained than the contents really are. The contents represent change; the individuals come to think that they represent A-change and B-change respectively. Uniform but conditional Some people who accept that on balance their phenomenology does not have the content that the world contains A-properties, nevertheless say that there is something psychologically compelling about the idea that the A properties are something like a best deserver to be what our temporal phenomenology is about. I won t address the question of whether I think this is right I don t but I do think that such a phenomenon does occur in some domains. So in this domain I ll address what I take to be the best explanation of that sort of phenomenon, and argue that it it would provide no support to ESP were it true in the domain of temporal phenomenology. The domain where I think it definitely holds is the content of the concept of qualitative experience itself. About that area I (Braddon-Mitchell 2003) and some others including John Hawthorne (Hawthorne 2003) argued that the concept has a conditional structure; the content of the concept qualia is something like If there are non-physical properties with which we are directly acquainted in experience these are qualia and necessarily so; otherwise qualia are certain functional properties. The ides was to account for the priority that dualism has. If in the actual world there are both dualistic properties and (physical) functional properties, then qualia are the dualistic ones. But if there are only the physical ones, then we should not be eliminativists. Rather we should accept that the qualia are the physical functional properties.
Against the illusion theory of temporal phenomenology 217 This was an account of the content of a concept; but equally it could be taken to be an account of the content of phenomenology. So in the case of temporal phenomenology the idea would be that the content of the phenomenology would be A-properties if there are any, but if not then B-properties. In the absence of any A-properties, on such a proposal, the phenomenology would represent certain B-properties (such as B-change). This would explain the intuition of priority that some claim to have. If we have no idea whether there are any A properties or not, one thing we would know is that our phenomenology represents A-properties if there are any. We do not know this about B properties; our phenomenology on this proposal represents B properties only if there are no A-Properties. But does this provide any support for ESP? Does it mean that we should infer that our phenomenology represents that there are A-properties? No. For on this proposal the accessible a priori content of our temporal phenomenology is merely conditional, and does not distinguish between the world containing or not containing A-properties. Of course the c-intension itself is more precise. It is a content of A-properties if there are any, of B-properties if there are only those. But what its content is depends on what properties there are. And thus the content of the phenomenology provides not guide to what properties there are: for to find out the content of the phenomenology, we must first find out by some independent means (perhaps metaphysical argument, or paying attention to science or both) what kinds of properties there in fact are. So even if there is some kind of priority to the idea of our phenomenology representing A-properties, on the only account of how this priority works, it gives no comfort to those who believe in ESP. Non-uniform intrinsic representation via representationalism and cognitive roles In the last couple of sections,we were assuming that there was indeed something like representational power to temporal phenomenology, but that it was abstract or conditional and thus didn t support ESP. Of course there is another possibility: that there is precise content, but that that content is not constant from person to person. This would be the most straightforward explanation of API.
218 David Braddon-Mitchell So here the idea is that we have different phenomenologies. Those who sincerely claim to have temporal phenomenology which represents the world as contain A-properties mostly do; and those who sincerely deny it most possess phenomenology which represents differently. If this is the case, then ESP is in just as much trouble. Because even if the content of phenomenology is a good guide to the world, in this case someone is mistaken: someone has illusory phenomenology. It s not clear how to tell who this is aside from settling the issue of what kind of properties the world in fact contains. Of course there might be ways of determining that someone is subject to perceptual illusion independently. I do not see how that would work in this case, but clearly more argument to show that is the case would be desirable. How might this be? Consider a version of intentionalism about phenomenology. According to one version of it, phenomenal properties are identical to, or perhaps logically supervene on, representational properties, and representational properties are functional. (consider the exploration of this view for the case of colour in Jackson 2007) What functional roles are involved is controversial, but suppose that in this case the roles are largely inferential. They are to do with our tendency to infer and judge. So someone who has the patterns of inference and judgement of a committed A-theorist will instantiate one set of functional roles, and someone who has the patterns of inference and judgement of a committed B-theorist will instantiate another. Since phenomenology supervenes on these representations which in turn supervene on these roles, it will turn out that the committed A-theorist because she is a committed A- theorist will have phenomenologies of perception which do indeed represent the world as containing A-properties. But equally, the committed B-theorist, because she is a committed B-theorist. I use the term committed in a particular way. I use it to mean actually instantiating in an unreflective way the dispositions to infer and judge that are characteristic of the theory. Of course it is possible that one could be an official B-theorist (intellectually you are convinced it s the best) while still actually instantiating the dispositions associated with the other theory. Thus this account doesn t necessarily predict that the phenomenology will track official philosophical positions. But it might predict that it will track hardened, committed views that have been held a long time and have influenced rapid judgement. What impact would such an account have on ESP? Yet again it undermines it; because
Against the illusion theory of temporal phenomenology 219 rather than phenomenology being a good guide to what theory we should believe, the phenomenology is determined by what theory we believe. And as we believe different theories, and have different phenomenologies, we have to settle those differences by other means. Broad or secondary content of phenomenology I hinted in the section on conditional content that it was consistent with the content being conditional, that there is in fact a uniform broad content which is the secondary intention of the phenomenal state. Many accounts of content will allow this. Causal theories of content might say that the content of the relevant phenomenologies are given by the worldly properties that cause it. Intentionalism and functionalism about phenomenology, of the kind discussed in the previous section, will say that the phenomenology itself is determined by the representational states, and those are in turn functional, and depend on the states in the world that the phenomenology interacts with (in the previous section I was considering a version of the view in which there s no difference in functional role in an A-world and a B-world). On both of these kinds of views, there is (assuming that there are no major functional and behavioural differences between A-theorists and B-theorists of a broad kind) likely to be a uniform broad content to our temporal phenomenology if there is any. If there are A-properties in the world, it ll be of those A-properties; if only B-properties it ll be of the B-properties (the twist that turns this into the conditional view is that if there are both A- and B- properties, the content is of the A-properties). So there might be broad representations of our phenomenology. And on this view what their content is is practically decisive about what properties there are. But, of course, the direction of dependence and the epistemological direction are both wrong. The content depends on the properties, but there is no obvious way to know what the content of the phenomenology is except by finding out what properties there are in the world. Exactly because the content is broad, it s not accessible except by empirical or metaphysical investigation. The phenomenology might be just what you get when you have certain representation, but you can t tell introspectively from the phenomenology what that representation is.
220 David Braddon-Mitchell Broad uniform content via naïve realism There is another route to a similar conclusion to the one in the previous section via another widespread view in the philosophy of perception. Suppose you hold a kind of naïve realism about phenomenology and representational states, of the kind you find in Mike Martin. (Martin 1997) According to views of this ilk, the represented states are literally part of the represented states. Representation is not solely in the head, but includes the parts of the world that are represented. For a view like this, content is entirely intrinsic to representational states for the simple reason that the wordly things are the content are literally parts of the state. Such views have always had a problem is with illusion, for on this kind of view there is little or nothing in common between a veridical state and an illusory state which makes someone judge that they are in a veridical state of the same content. Here s a very simple example: At one time Fred has a veridical phenomenal state as of a red apple, and later he claims (as the result of an hallucination) to be in the same state, but in fact there is no apple and he is hallucinating. On most traditional views he might have been in the same phenomenological state on both occasions, which had the same content on both occasions, but for some reason it was improperly caused the second time round so its content was false: it was a state as of an apple, but there was no apple. The naïve realist cannot help herself to this account. For according to her the two states have completely different contents, for the content of the state in the first case is literally the apple, but in the second there is no apple. And, on this view, the phenomenology goes with the content so they must be quite different phenomenologies. Something like this is the view known as disjunctivism. And one thing the disjunctivist has to accept, it would seem, is that there is more to phenomenologies that we can tell. In other words, it is possible to be mistaken about having the same phenomenology. For in the Fred case, if he believes that he is in the same state both times it is not just that he is mistaken about there being an apple, he is also mistaken about what his phenomenology is. Perhaps the reader has seen where this is going. The disjunctivist, if right, has to accept that we are in no position to tell what the content of our phenomenology is by inward means. You find out what your thoughts are about by being directed towards the world finding out whether there are apples connected to you in the right way, and under what circumstances you can be mistaken about that. By discovering that you are in one of those situations in
Against the illusion theory of temporal phenomenology 221 which there are no apples though you are inclined to judge phenomenally that there are, you find out that your phenomenology is not of apples after all. This has considerable bearing on the phenomenology of time. Is the A-theorists phenomenology one which represents the world as containing A-properties? This is not something that is settled internally. You settle in by looking to the world and trying to establish if there are A-properties that could be part of your experience. In other words again you establish it by metaphysics, science or both. Yet again it s true on this view that if the content of our experience is on A-properties, A-properties exist. But, also again, there is no way to establish what the content of our experience is without independently establishing whether there are A-properties. Conclusion None of this of course establishes that ESP is false; it just establishes that on a very large and representative sample of views about the phenomenology of content it must be. This then narrows the size of logical space the truth of ESP can occupy. It s the fragment of logical space in which we accept that we have phenomenal states which have introspectible content which intrinsically, unambiguously, unconditionally, uniformly and reliably presents the world to us in a way which give us reason to believe in A-properties. I m not even sure how such a combination would work; perhaps via a version of the doctrine of revelation for qualia which goes beyond the nature of the qualia themselves and on to what they represent. Defending such a combination is, I think, a big ask. It has been boxed into a very small box indeed. Importantly this logical box has to have space not just for the A-theorist who gets there by (or is motivated by) ESP, but also the B-theoretic illusion theorist who accepts ESP but relies on a special explanation of why the phenomenology is misleading (perhaps this latter position can, as it were, poke its elbow into a neighbouring piece of logical space in which the reliability constraint is relaxed a little.) And even if there were powerful arguments for this conjunction of views, they would still have trouble explaining the data: for if we had states like this, how could it be that the present author, and many others, seem sincerely to deny that they have such phenomenology.
222 David Braddon-Mitchell References [1] Braddon-Mitchell, D. (2003) Qualia and Analytical Conditionals, The Journal of Philosophy 100(3): 111 135. [2] Hawthorne, J. (2002) Advice to Physicalists, Philosophical Studies 109(1): 17 52. [3] Jackson, F. (2007) Colour for Representationalists, Erkenntnis 66(1-2): 169 85. [4] Martin, M. G. F. (1997) The Reality of Appearances in M. Sainsbury (ed.), Thought and Ontology, Milan: FrancoAngeli, 81 106. [5] Paul, L. (2010) Temporal Experience, Journal of Philosophy 107(7): 333. [6] Prosser, S. (2013) Passage and Perception, Noûs 47(1): 69 84. [7] Skow, B. (2011) Experience and the Passage of TIme, Philosophical Perspectives 25 (1): 359 387. [8] Smart, J. J. C. (2006) Metaphysical Illusions, Australasian Journal of Philosophy 84(2): 167 175.